Speed improvement in jnt_convolve y and 2d_vert for 32 bit Neon Observed 40% gain in jnt_convolve_y_neon and 27% gain in jnt_convolve_2d_vert_neon w.r.t. existing neon code in unit test. Change-Id: Ia5a9095ea78337d863c8aa43876fc93428385d85
diff --git a/av1/common/arm/jnt_convolve_neon.c b/av1/common/arm/jnt_convolve_neon.c index 4015082..0bc9d92 100644 --- a/av1/common/arm/jnt_convolve_neon.c +++ b/av1/common/arm/jnt_convolve_neon.c
@@ -22,12 +22,106 @@ #include "av1/common/arm/mem_neon.h" #include "av1/common/arm/transpose_neon.h" +static INLINE void compute_avg_4x1(uint16x4_t res0, uint16x4_t d0, + const uint16_t fwd_offset, + const uint16_t bck_offset, + const int16x4_t sub_const_vec, + const int16_t round_bits, + const int use_jnt_comp_avg, uint8x8_t *t0) { + int16x4_t tmp0; + uint16x4_t tmp_u0; + uint32x4_t sum0; + int32x4_t dst0; + int16x8_t tmp4; + + if (use_jnt_comp_avg) { + const int32x4_t round_bits_vec = vdupq_n_s32((int32_t)(-round_bits)); + + sum0 = vmull_n_u16(res0, fwd_offset); + sum0 = vmlal_n_u16(sum0, d0, bck_offset); + + sum0 = vshrq_n_u32(sum0, DIST_PRECISION_BITS); + + dst0 = vsubq_s32(vreinterpretq_s32_u32(sum0), vmovl_s16(sub_const_vec)); + + dst0 = vqrshlq_s32(dst0, round_bits_vec); + + tmp0 = vqmovn_s32(dst0); + tmp4 = vcombine_s16(tmp0, tmp0); + + *t0 = vqmovun_s16(tmp4); + } else { + const int16x4_t round_bits_vec = vdup_n_s16(-round_bits); + tmp_u0 = vhadd_u16(res0, d0); + + tmp0 = vsub_s16(vreinterpret_s16_u16(tmp_u0), sub_const_vec); + + tmp0 = vqrshl_s16(tmp0, round_bits_vec); + + tmp4 = vcombine_s16(tmp0, tmp0); + + *t0 = vqmovun_s16(tmp4); + } +} + +static INLINE void compute_avg_8x1(uint16x8_t res0, uint16x8_t d0, + const uint16_t fwd_offset, + const uint16_t bck_offset, + const int16x4_t sub_const, + const int16_t round_bits, + const int use_jnt_comp_avg, uint8x8_t *t0) { + int16x4_t tmp0, tmp2; + int16x8_t f0; + uint32x4_t sum0, sum2; + int32x4_t dst0, dst2; + + uint16x8_t tmp_u0; + + if (use_jnt_comp_avg) { + const int32x4_t sub_const_vec = vmovl_s16(sub_const); + const int32x4_t round_bits_vec = vdupq_n_s32(-(int32_t)round_bits); + + sum0 = vmull_n_u16(vget_low_u16(res0), fwd_offset); + sum0 = vmlal_n_u16(sum0, vget_low_u16(d0), bck_offset); + sum0 = vshrq_n_u32(sum0, DIST_PRECISION_BITS); + + sum2 = vmull_n_u16(vget_high_u16(res0), fwd_offset); + sum2 = vmlal_n_u16(sum2, vget_high_u16(d0), bck_offset); + sum2 = vshrq_n_u32(sum2, DIST_PRECISION_BITS); + + dst0 = vsubq_s32(vreinterpretq_s32_u32(sum0), sub_const_vec); + dst2 = vsubq_s32(vreinterpretq_s32_u32(sum2), sub_const_vec); + + dst0 = vqrshlq_s32(dst0, round_bits_vec); + dst2 = vqrshlq_s32(dst2, round_bits_vec); + + tmp0 = vqmovn_s32(dst0); + tmp2 = vqmovn_s32(dst2); + + f0 = vcombine_s16(tmp0, tmp2); + + *t0 = vqmovun_s16(f0); + + } else { + const int16x8_t sub_const_vec = vcombine_s16(sub_const, sub_const); + const int16x8_t round_bits_vec = vdupq_n_s16(-round_bits); + + tmp_u0 = vhaddq_u16(res0, d0); + + f0 = vsubq_s16(vreinterpretq_s16_u16(tmp_u0), sub_const_vec); + + f0 = vqrshlq_s16(f0, round_bits_vec); + + *t0 = vqmovun_s16(f0); + } +} + static INLINE void compute_avg_4x4( uint16x4_t res0, uint16x4_t res1, uint16x4_t res2, uint16x4_t res3, uint16x4_t d0, uint16x4_t d1, uint16x4_t d2, uint16x4_t d3, const uint16_t fwd_offset, const uint16_t bck_offset, const int16x4_t sub_const_vec, const int16_t round_bits, - const int32_t use_jnt_comp_avg, uint8x8_t *t0, uint8x8_t *t1) { + const int use_jnt_comp_avg, uint8x8_t *t0, uint8x8_t *t1) { int16x4_t tmp0, tmp1, tmp2, tmp3; uint16x4_t tmp_u0, tmp_u1, tmp_u2, tmp_u3; uint32x4_t sum0, sum1, sum2, sum3; @@ -107,7 +201,7 @@ uint16x8_t d0, uint16x8_t d1, uint16x8_t d2, uint16x8_t d3, const uint16_t fwd_offset, const uint16_t bck_offset, const int16x4_t sub_const, const int16_t round_bits, - const int32_t use_jnt_comp_avg, uint8x8_t *t0, uint8x8_t *t1, uint8x8_t *t2, + const int use_jnt_comp_avg, uint8x8_t *t0, uint8x8_t *t1, uint8x8_t *t2, uint8x8_t *t3) { int16x4_t tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7; int16x8_t f0, f1, f2, f3; @@ -231,7 +325,6 @@ int16_t *dst_ptr; int dst_stride; int width, height; - uint8x8_t t0, t1, t2, t3, t4, t5, t6, t7; dst_ptr = im_block; dst_stride = im_stride; @@ -239,15 +332,22 @@ width = w; if (w == 4) { - int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10, d0, d1, d2, d3; - int16x8_t tt0, tt1, tt2, tt3; + int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, d0; + int16x8_t tt0; + uint8x8_t t0; const int16x4_t horiz_const = vdup_n_s16((1 << (bd + FILTER_BITS - 2))); const int16x4_t shift_round_0 = vdup_n_s16(-(round_0)); +#if defined(__aarch64__) + int16x4_t s8, s9, s10, d1, d2, d3; + int16x8_t tt1, tt2, tt3; + uint8x8_t t1, t2, t3; +#endif do { s = src; __builtin_prefetch(s + 0 * src_stride); +#if defined(__aarch64__) __builtin_prefetch(s + 1 * src_stride); __builtin_prefetch(s + 2 * src_stride); __builtin_prefetch(s + 3 * src_stride); @@ -301,17 +401,48 @@ src += 4 * src_stride; dst_ptr += 4 * dst_stride; height -= 4; +#else + t0 = vld1_u8(s); // a0 a1 a2 a3 a4 a5 a6 a7 + tt0 = vreinterpretq_s16_u16(vmovl_u8(t0)); // a0 a1 a2 a3 a4 a5 a6 a7 + s0 = vget_low_s16(tt0); // a0 a1 a2 a3 + s4 = vget_high_s16(tt0); // a4 a5 a6 a7 + __builtin_prefetch(dst_ptr); + s += 8; + t0 = vld1_u8(s); // a8 a9 a10 a11 + + // a8 a9 a10 a11 + s7 = vget_low_s16(vreinterpretq_s16_u16(vmovl_u8(t0))); + + s1 = vext_s16(s0, s4, 1); // a1 a2 a3 a4 + s2 = vext_s16(s0, s4, 2); // a2 a3 a4 a5 + s3 = vext_s16(s0, s4, 3); // a3 a4 a5 a6 + s5 = vext_s16(s4, s7, 1); // a5 a6 a7 a8 + s6 = vext_s16(s4, s7, 2); // a6 a7 a8 a9 + s7 = vext_s16(s4, s7, 3); // a7 a8 a9 a10 + + d0 = convolve8_4x4_s16(s0, s1, s2, s3, s4, s5, s6, s7, x_filter_tmp, + horiz_const, shift_round_0); + + vst1_s16(dst_ptr, d0); + + src += src_stride; + dst_ptr += dst_stride; + height -= 1; +#endif } while (height > 0); } else { int16_t *d_tmp; - int16x8_t s11, s12, s13, s14; - int16x8_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10; - int16x8_t res0, res1, res2, res3, res4, res5, res6, res7; + int16x8_t s0, s1, s2, s3, s4, s5, s6, s7; + int16x8_t res0; + uint8x8_t t0; const int16x8_t horiz_const = vdupq_n_s16((1 << (bd + FILTER_BITS - 2))); const int16x8_t shift_round_0 = vdupq_n_s16(-(round_0)); - do { +#if defined(__aarch64__) + uint8x8_t t1, t2, t3, t4, t5, t6, t7; + int16x8_t s8, s9, s10, s11, s12, s13, s14; + int16x8_t res1, res2, res3, res4, res5, res6, res7; __builtin_prefetch(src + 0 * src_stride); __builtin_prefetch(src + 1 * src_stride); __builtin_prefetch(src + 2 * src_stride); @@ -390,6 +521,42 @@ src += 8 * src_stride; dst_ptr += 8 * dst_stride; height -= 8; +#else + int16x8_t temp_0; + t0 = vld1_u8(src); + s0 = vreinterpretq_s16_u16(vmovl_u8(t0)); // a0 a1 a2 a3 a4 a5 a6 a7 + + width = w; + s = src + 8; + d_tmp = dst_ptr; + __builtin_prefetch(dst_ptr); + + do { + t0 = vld1_u8(s); // a8 a9 a10 a11 a12 a13 a14 a15 + s7 = vreinterpretq_s16_u16(vmovl_u8(t0)); + temp_0 = s0; + s0 = s7; + + s1 = vextq_s16(temp_0, s7, 1); // a1 a2 a3 a4 a5 a6 a7 a8 + s2 = vextq_s16(temp_0, s7, 2); // a2 a3 a4 a5 a6 a7 a8 a9 + s3 = vextq_s16(temp_0, s7, 3); // a3 a4 a5 a6 a7 a8 a9 a10 + s4 = vextq_s16(temp_0, s7, 4); // a4 a5 a6 a7 a8 a9 a10 a11 + s5 = vextq_s16(temp_0, s7, 5); // a5 a6 a7 a8 a9 a10 a11 a12 + s6 = vextq_s16(temp_0, s7, 6); // a6 a7 a8 a9 a10 a11 a12 a13 + s7 = vextq_s16(temp_0, s7, 7); // a7 a8 a9 a10 a11 a12 a13 a14 + + res0 = convolve8_8x8_s16(temp_0, s1, s2, s3, s4, s5, s6, s7, + x_filter_tmp, horiz_const, shift_round_0); + vst1q_s16(d_tmp, res0); + + s += 8; + d_tmp += 8; + width -= 8; + } while (width > 0); + src += src_stride; + dst_ptr += dst_stride; + height -= 1; +#endif } while (height > 0); } } @@ -420,10 +587,15 @@ const int do_average = conv_params->do_average; const int use_jnt_comp_avg = conv_params->use_jnt_comp_avg; - int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10; - uint16x4_t res4, res5, res6, res7; - uint16x4_t d0, d1, d2, d3; - uint8x8_t t0, t1; + int16x4_t s0, s1, s2, s3, s4, s5, s6, s7; + uint16x4_t res4, d0; + uint8x8_t t0; + +#if defined(__aarch64__) + int16x4_t s8, s9, s10; + uint16x4_t res5, res6, res7, d1, d2, d3; + uint8x8_t t1; +#endif dst = conv_params->dst; src_ptr = im_block; @@ -450,6 +622,7 @@ s += (7 * im_stride); do { +#if defined(__aarch64__) load_s16_4x4(s, im_stride, &s7, &s8, &s9, &s10); s += (im_stride << 2); @@ -480,17 +653,13 @@ bck_offset, sub_const_vec, round_bits, use_jnt_comp_avg, &t0, &t1); - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), - 0); // 00 01 02 03 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), 0); d_u8 += dst8_stride; - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), - 1); // 10 11 12 13 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), 1); d_u8 += dst8_stride; - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), - 0); // 20 21 22 23 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), 0); d_u8 += dst8_stride; - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), - 1); // 30 31 32 33 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), 1); d_u8 += dst8_stride; } else { @@ -505,6 +674,39 @@ s5 = s9; s6 = s10; height -= 4; +#else + s7 = vld1_s16(s); + s += (im_stride); + + __builtin_prefetch(d + 0 * dst_stride); + __builtin_prefetch(d_u8 + 0 * dst8_stride); + + d0 = convolve8_4x4_s32(s0, s1, s2, s3, s4, s5, s6, s7, y_filter, + round_shift_vec, offset_const); + + if (do_average) { + res4 = vld1_u16(d); + d += (dst_stride); + + compute_avg_4x1(res4, d0, fwd_offset, bck_offset, sub_const_vec, + round_bits, use_jnt_comp_avg, &t0); + + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), 0); + d_u8 += dst8_stride; + + } else { + vst1_u16(d, d0); + d += (dst_stride); + } + s0 = s1; + s1 = s2; + s2 = s3; + s3 = s4; + s4 = s5; + s5 = s6; + s6 = s7; + height--; +#endif } while (height > 0); src_ptr += 4; dst_ptr += 4; @@ -722,8 +924,10 @@ uint8_t *dst_u8_ptr; CONV_BUF_TYPE *d, *dst_ptr; int width, height; - uint8x8_t t0, t1, t2, t3, t4, t5, t6, t7; - + uint8x8_t t0; +#if defined(__aarch64__) + uint8x8_t t1, t2, t3, t4, t5, t6, t7; +#endif s = src_ptr; dst_ptr = dst; dst_u8_ptr = dst8; @@ -731,11 +935,18 @@ height = h; if ((w == 4) || (h == 4)) { - int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10, d0, d1, d2, d3; - int16x8_t tt0, tt1, tt2, tt3; - uint16x4_t res4, res5, res6, res7; + int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, d0; + int16x8_t tt0; + uint16x4_t res4; +#if defined(__aarch64__) + int16x4_t s8, s9, s10, d1, d2, d3; + int16x8_t tt1, tt2, tt3; + uint16x4_t res5, res6, res7; uint32x2_t tu0, tu1; int16x8_t u0, u1; +#else + int16x4_t temp_0; +#endif const int16x4_t zero = vdup_n_s16(0); const int16x4_t round_offset_vec = vdup_n_s16(round_offset); const int16x4_t shift_round_0 = vdup_n_s16(-conv_params->round_0 + 1); @@ -746,6 +957,7 @@ d_u8 = dst_u8_ptr; width = w; __builtin_prefetch(s + 0 * src_stride); +#if defined(__aarch64__) __builtin_prefetch(s + 1 * src_stride); __builtin_prefetch(s + 2 * src_stride); __builtin_prefetch(s + 3 * src_stride); @@ -854,15 +1066,66 @@ dst_ptr += (dst_stride << 2); dst_u8_ptr += (dst8_stride << 2); height -= 4; +#else + t0 = vld1_u8(s); // a0 a1 a2 a3 a4 a5 a6 a7 + tt0 = vreinterpretq_s16_u16(vmovl_u8(t0)); // a0 a1 a2 a3 a4 a5 a6 a7 + s0 = vget_low_s16(tt0); // a0 a1 a2 a3 + s4 = vget_high_s16(tt0); // a4 a5 a6 a7 + __builtin_prefetch(d); + + s += 8; + do { + t0 = vld1_u8(s); // a8 a9 a10 a11 + + // a8 a9 a10 a11 + s7 = vget_low_s16(vreinterpretq_s16_u16(vmovl_u8(t0))); + temp_0 = s7; + s1 = vext_s16(s0, s4, 1); // a1 a2 a3 a4 + s2 = vext_s16(s0, s4, 2); // a2 a3 a4 a5 + s3 = vext_s16(s0, s4, 3); // a3 a4 a5 a6 + s5 = vext_s16(s4, s7, 1); // a5 a6 a7 a8 + s6 = vext_s16(s4, s7, 2); // a6 a7 a8 a9 + s7 = vext_s16(s4, s7, 3); // a7 a8 a9 a10 + + d0 = convolve8_4x4_s16(s0, s1, s2, s3, s4, s5, s6, s7, x_filter_tmp, + zero, shift_round_0); + d0 = vrshl_s16(d0, horiz_const); + d0 = vadd_s16(d0, round_offset_vec); + s0 = s4; + s4 = temp_0; + if (conv_params->do_average) { + __builtin_prefetch(d); + __builtin_prefetch(d_u8); + + res4 = vld1_u16(d); + + compute_avg_4x1(res4, vreinterpret_u16_s16(d0), fwd_offset, + bck_offset, round_offset_vec, round_bits, + use_jnt_comp_avg, &t0); + + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), + 0); // 00 01 02 03 + } else { + vst1_u16(d, vreinterpret_u16_s16(d0)); + } + + s += 4; + width -= 4; + d += 4; + d_u8 += 4; + } while (width > 0); + src_ptr += (src_stride); + dst_ptr += (dst_stride); + dst_u8_ptr += (dst8_stride); + height--; +#endif } while (height > 0); } else { CONV_BUF_TYPE *d_tmp; uint8_t *d_u8_tmp; - int16x8_t s11, s12, s13, s14; - int16x8_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10; - int16x8_t res0, res1, res2, res3, res4, res5, res6, res7; - uint16x8_t res8, res9, res10, res11; - + int16x8_t s0, s1, s2, s3, s4, s5, s6, s7; + int16x8_t res0; + uint16x8_t res8; const int16x8_t round_offset128 = vdupq_n_s16(round_offset); const int16x4_t round_offset64 = vdup_n_s16(round_offset); const int16x8_t shift_round_0 = vdupq_n_s16(-conv_params->round_0 + 1); @@ -872,6 +1135,11 @@ d = dst_ptr = dst; d_u8 = dst_u8_ptr = dst8; do { +#if defined(__aarch64__) + int16x8_t s11, s12, s13, s14; + int16x8_t s8, s9, s10; + int16x8_t res1, res2, res3, res4, res5, res6, res7; + uint16x8_t res9, res10, res11; __builtin_prefetch(src_ptr + 0 * src_stride); __builtin_prefetch(src_ptr + 1 * src_stride); __builtin_prefetch(src_ptr + 2 * src_stride); @@ -1007,6 +1275,67 @@ dst_ptr += 8 * dst_stride; dst_u8_ptr += 8 * dst8_stride; height -= 8; +#else + int16x8_t temp_0; + __builtin_prefetch(src_ptr); + t0 = vld1_u8(src_ptr); + s0 = vreinterpretq_s16_u16(vmovl_u8(t0)); // a0 a1 a2 a3 a4 a5 a6 a7 + + width = w; + s = src_ptr + 8; + d = dst_ptr; + d_u8_tmp = dst_u8_ptr; + + __builtin_prefetch(dst_ptr); + + do { + d_u8 = d_u8_tmp; + d_tmp = d; + + t0 = vld1_u8(s); // a8 a9 a10 a11 a12 a13 a14 a15 + s7 = vreinterpretq_s16_u16(vmovl_u8(t0)); + temp_0 = s0; + s0 = s7; + + s1 = vextq_s16(temp_0, s7, 1); // a1 a2 a3 a4 a5 a6 a7 a8 + s2 = vextq_s16(temp_0, s7, 2); // a2 a3 a4 a5 a6 a7 a8 a9 + s3 = vextq_s16(temp_0, s7, 3); // a3 a4 a5 a6 a7 a8 a9 a10 + s4 = vextq_s16(temp_0, s7, 4); // a4 a5 a6 a7 a8 a9 a10 a11 + s5 = vextq_s16(temp_0, s7, 5); // a5 a6 a7 a8 a9 a10 a11 a12 + s6 = vextq_s16(temp_0, s7, 6); // a6 a7 a8 a9 a10 a11 a12 a13 + s7 = vextq_s16(temp_0, s7, 7); // a7 a8 a9 a10 a11 a12 a13 a14 + + res0 = convolve8_8x8_s16(temp_0, s1, s2, s3, s4, s5, s6, s7, + x_filter_tmp, zero, shift_round_0); + + res0 = vrshlq_s16(res0, horiz_const); + res0 = vaddq_s16(res0, round_offset128); + + if (conv_params->do_average) { + res8 = vld1q_u16(d_tmp); + d_tmp += (dst_stride); + + compute_avg_8x1(res8, vreinterpretq_u16_s16(res0), fwd_offset, + bck_offset, round_offset64, round_bits, + use_jnt_comp_avg, &t0); + + vst1_u8(d_u8, t0); + d_u8 += (dst8_stride); + } else { + vst1q_u16(d_tmp, vreinterpretq_u16_s16(res0)); + d_tmp += (dst_stride); + } + + s += 8; + d += 8; + width -= 8; + d_u8_tmp += 8; + } while (width > 0); + src_ptr += src_stride; + dst_ptr += dst_stride; + dst_u8_ptr += dst8_stride; + height--; +#endif } while (height > 0); } } @@ -1057,7 +1386,6 @@ uint8_t *dst_u8_ptr; CONV_BUF_TYPE *d, *dst_ptr; int width, height; - uint8x8_t t0, t1, t2, t3, t4, t5, t6, t7; s = src_ptr; dst_ptr = dst; @@ -1070,11 +1398,17 @@ assert((conv_params->round_1 - 2) >= bits); if ((w == 4) || (h == 4)) { - int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10, d0, d1, d2, d3; - uint16x4_t res4, res5, res6, res7; + int16x4_t s0, s1, s2, s3, s4, s5, s6, s7, d0; + uint16x4_t res4; uint32x2_t tu0, tu1, tu2, tu3; int16x8_t u0, u1, u2, u3; + uint8x8_t t0; +#if defined(__aarch64__) + int16x4_t s8, s9, s10, d1, d2, d3; + uint16x4_t res5, res6, res7; + uint8x8_t t1; +#endif const int16x4_t round_offset64 = vdup_n_s16(round_offset); const int16x4_t shift_vec = vdup_n_s16(-shift_value); const int16x4_t zero = vdup_n_s16(0); @@ -1111,6 +1445,7 @@ s += (7 * src_stride); do { +#if defined(__aarch64__) load_unaligned_u8_4x4(s, src_stride, &tu0, &tu1); u0 = vreinterpretq_s16_u16(vmovl_u8(vreinterpret_u8_u32(tu0))); @@ -1154,17 +1489,13 @@ round_offset64, round_bits, use_jnt_comp_avg, &t0, &t1); - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), - 0); // 00 01 02 03 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), 0); d_u8 += dst8_stride; - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), - 1); // 10 11 12 13 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), 1); d_u8 += dst8_stride; - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), - 0); // 20 21 22 23 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), 0); d_u8 += dst8_stride; - vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), - 1); // 30 31 32 33 + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t1), 1); d_u8 += dst8_stride; } else { store_u16_4x4(d, dst_stride, vreinterpret_u16_s16(d0), @@ -1183,6 +1514,44 @@ s += (src_stride << 2); height -= 4; +#else + load_unaligned_u8_4x1(s, src_stride, &tu0); + u0 = vreinterpretq_s16_u16(vmovl_u8(vreinterpret_u8_u32(tu0))); + s7 = vget_low_s16(u0); + + d0 = convolve8_4x4_s16(s0, s1, s2, s3, s4, s5, s6, s7, y_filter_tmp, + zero, shift_vec); + + d0 = vadd_s16(d0, round_offset64); + + if (conv_params->do_average) { + __builtin_prefetch(d); + + res4 = vld1_u16(d); + d += (dst_stride); + + compute_avg_4x1(res4, vreinterpret_u16_s16(d0), fwd_offset, + bck_offset, round_offset64, round_bits, + use_jnt_comp_avg, &t0); + + vst1_lane_u32((uint32_t *)d_u8, vreinterpret_u32_u8(t0), 0); + d_u8 += dst8_stride; + } else { + vst1_u16(d, vreinterpret_u16_s16(d0)); + d += (dst_stride); + } + + s0 = s1; + s1 = s2; + s2 = s3; + s3 = s4; + s4 = s5; + s5 = s6; + s6 = s7; + + s += (src_stride); + height--; +#endif } while (height > 0); src_ptr += 4; dst_ptr += 4; @@ -1191,15 +1560,19 @@ } while (width > 0); } else { CONV_BUF_TYPE *d_tmp; - int16x8_t s11, s12, s13, s14; - int16x8_t s0, s1, s2, s3, s4, s5, s6, s7, s8, s9, s10; - int16x8_t res0, res1, res2, res3, res4, res5, res6, res7; - uint16x8_t res8, res9, res10, res11; + int16x8_t s0, s1, s2, s3, s4, s5, s6, s7; + int16x8_t res0; + uint16x8_t res8; + uint8x8_t t0, t1, t2, t3, t4, t5, t6, t7; const int16x8_t round_offset128 = vdupq_n_s16(round_offset); const int16x8_t shift_vec = vdupq_n_s16(-shift_value); const int16x4_t round_offset64 = vdup_n_s16(round_offset); const int16x8_t zero = vdupq_n_s16(0); - +#if defined(__aarch64__) + int16x8_t s8, s9, s10, s11, s12, s13, s14; + int16x8_t res1, res2, res3, res4, res5, res6, res7; + uint16x8_t res10, res11, res9; +#endif dst_ptr = dst; dst_u8_ptr = dst8; do { @@ -1227,6 +1600,7 @@ d_u8 = dst_u8_ptr; do { +#if defined(__aarch64__) load_u8_8x8(s, src_stride, &t0, &t1, &t2, &t3, &t4, &t5, &t6, &t7); s7 = vreinterpretq_s16_u16(vmovl_u8(t0)); @@ -1316,6 +1690,43 @@ s6 = s14; s += (8 * src_stride); height -= 8; +#else + s7 = vreinterpretq_s16_u16(vmovl_u8(vld1_u8(s))); + + __builtin_prefetch(dst_ptr); + + res0 = convolve8_8x8_s16(s0, s1, s2, s3, s4, s5, s6, s7, y_filter_tmp, + zero, shift_vec); + res0 = vaddq_s16(res0, round_offset128); + + s0 = s1; + s1 = s2; + s2 = s3; + s3 = s4; + s4 = s5; + s5 = s6; + s6 = s7; + + if (conv_params->do_average) { + __builtin_prefetch(d_tmp); + + res8 = vld1q_u16(d_tmp); + d_tmp += (dst_stride); + + compute_avg_8x1(res8, vreinterpretq_u16_s16(res0), fwd_offset, + bck_offset, round_offset64, round_bits, + use_jnt_comp_avg, &t0); + + vst1_u8(d_u8, t0); + d_u8 += (dst8_stride); + } else { + vst1q_u16(d_tmp, vreinterpretq_u16_s16(res0)); + d_tmp += dst_stride; + } + + s += (src_stride); + height--; +#endif } while (height > 0); src_ptr += 8; dst_ptr += 8;
diff --git a/av1/common/arm/mem_neon.h b/av1/common/arm/mem_neon.h index 4bf45a5..d5f458e 100644 --- a/av1/common/arm/mem_neon.h +++ b/av1/common/arm/mem_neon.h
@@ -362,6 +362,15 @@ *tu1 = vset_lane_u32(a, *tu1, 1); } +static INLINE void load_unaligned_u8_4x1(const uint8_t *buf, int stride, + uint32x2_t *tu0) { + uint32_t a; + + memcpy(&a, buf, 4); + buf += stride; + *tu0 = vset_lane_u32(a, *tu0, 0); +} + static INLINE void load_unaligned_u8_4x2(const uint8_t *buf, int stride, uint32x2_t *tu0) { uint32_t a;